ATMS SDR Overview

Fuzhong Weng

ATMS SDR Team August 24, 2015

STAR JPSS 2015 Annual Science Team Meeting August 24-28, 2015

5830 University Research Court, College Park, MD 20740

Outline

2

• ATMS SDR Team Members • ATMS Instrument Overview • ATMS SNPP Product Overview • ATMS JPSS-1 Readiness • Summary and Path Forward

ATMS SDR Team Members

3

PI Name Organization Primary Role and Responsibility

Fuzhong Weng NOAA Budget execution, calval task planning, and ATMS SDR sciences and algorithms

Ninghai Sun ERT Technical coordination, ATMS SDR processing in ADL, ATMS monitoring, anomaly investigation

Edward Kim NASA NASA ATMS instrument scientist, TVAC data, instrument anomaly investigation

Vince Leslie MIT/LL Calval support, SDR sciences, PCT/LUT, prelaunch TVAC data analysis, RDR generation

Xiaolei Zou ESSIC/UMD Striping analysis and mitigation, RFI analysis, xcal (ATMS vs. AMSU)

Kent Anderson NGES NGES ATMS instrument and engineering sciences, TVAC data

Wes Berg CIRA Xcal ATMS with GPM microwave imager, other WG band instruemnts

Wael Ibrahim Raytheon IDPS operational feedbacks and code implemention

Hu(Tiger) Yang

ESSIC/UMD

ATMS SDR algorithm sciences, full radiance calibration, lunar correction, antenna spill-over

ATMS Instrument Review

• SNPP and JPSS-1 ATMS instruments are identical in channel and spatial resolution. J1 ATMS design life should be longer than SNPP ATMS (e.g. better bearing system)

• SNPP ATMS has been commanded for daily scan reversal to extend the life time performance beyond 5 years. The motor current shows significant drops after the August 17 reversal.

– The reversal was initiated above 75 degrees latitude North and repeats every 15 orbits after the previous reversal. The 15 orbits will "walk" the longitude across the earth in 14-15 days, with steps about 20 degrees longitude between successive orbits. Each reversal last no more than 1 minute

• Rework of JPSS-1 ATMS is nearly completed. The new TVAC data will be soon released. The sensor will be delivered on November 7

– Team has evaluated the proposal of TVAC with less scene measurements from 11 to 6.

– All the software for anlayzing TVAC data is ready at STAR

4

1st reversal

2nd reversal

ATMS Instrument Characterization

5

Ch Channel Central

Freq.(MHz) Polarization

Bandwidth Max. (MHz)

Frequency Stability (MHz)

Calibration Accuracy

(K)

Nonlinearity Max. (K)

NEΔT (K)

3-dB Bandwidth

(deg) Remarks Characterization at Nadir

1 23800 QV 270 10 1.0 0.3 0.5 5.2 AMSU-A2 Window-water vapor 100 mm

2 31400 QV 180 10 1.0 0.4 0.6 5.2 AMSU-A2 Window-water vapor 500 mm

3 50300 QH 180 10 0.75 0.4 0.7 2.2 AMSU-A1-2 Window-surface emissivity

4 51760 QH 400 5 0.75 0.4 0.5 2.2 Window-surface emissivity

5 52800 QH 400 5 0.75 0.4 0.5 2.2 AMSU-A1-2 Surface air

6 53596±115 QH 170 5 0.75 0.4 0.5 2.2 AMSU-A1-2 4 km ~ 700 mb

7 54400 QH 400 5 0.75 0.4 0.5 2.2 AMSU-A1-1 9 km ~ 400 mb

8 54940 QH 400 10 0.75 0.4 0.5 2.2 AMSU-A1-1 11 km ~ 250 mb

9 55500 QH 330 10 0.75 0.4 0.5 2.2 AMSU-A1-2 13 km ~ 180 mb

10 57290.344(fo) QH 330 0.5 0.75 0.4 0.75 2.2 AMSU-A1-1 17 km ~ 90 mb

11 fo± 217 QH 78 0.5 0.75 0.4 1.0 2.2 AMSU-A1-1 19 km ~ 50 mb

12 fo±322.2±48 QH 36 1.2 0.75 0.4 1.0 2.2 AMSU-A1-1 25 km ~ 25 mb

13 fo±322.2±22 QH 16 1.6 0.75 0.4 1.5 2.2 AMSU-A1-1 29 km ~ 10 mb

14 fo±322.2±10 QH 8 0.5 0.75 0.4 2.2 2.2 AMSU-A1-1 32 km ~ 6 mb

15 fo±322.2±4.5 QH 3 0.5 0.75 0.4 3.6 2.2 AMSU-A1-1 37 km ~ 3 mb

16 88200 QV 2000 200 1.0 0.4 0.3 2.2 89000 Window H2O 150 mm

17 165500 QH 3000 200 1.0 0.4 0.6 1.1 157000 H2O 18 mm

18 183310±7000 QH 2000 30 1.0 0.4 0.8 1.1 AMSU-B H2O 8 mm

19 183310±4500 QH 2000 30 1.0 0.4 0.8 1.1 H2O 4.5 mm

20 183310±3000 QH 1000 30 1.0 0.4 0.8 1.1 AMSU-B/MHS H2O 2.5 mm

21 183310±1800 QH 1000 30 1.0 0.4 0.8 1.1 H2O 1.2 mm

22 183310±1000 QH 500 30 1.0 0.4 0.9 1.1 AMSU-B/MHS H2O 0.5 mm

Ch GHz Pol Ch GHz Pol Ch GHz Pol

1 23.8 QV 1 23.8 QV

2 31.399 QV 2 31.4 QV

1 50.299 QV 3 50.299 QV 3 50.3 QH

4 51.76 QH

4 52.8 QV 5 52.8 QH

2 53.74 QH 5 53.595 ± 0.115 QH 6 53.596 ± 0.115 QH

6 54.4 QH 7 54.4 QH

3 54.96 QH 7 54.94 QV 8 54.94 QH

8 55.5 QH 9 55.5 QH

4 57.95 QH 9 fo = 57.29 QH 10 fo = 57.29 QH

10 fo ± 0.217 QH 11 fo±0.3222±0.217 QH

11 fo±0.3222±0.048 QH 12 fo± 0.3222±0.048 QH

12 fo ±0.3222±0.022 QH 13 fo±0.3222±0.022 QH

13 fo± 0.3222±0.010 QH 14 fo±0.3222 ±0.010 QH

14 fo±0.3222±0.0045 QH 15 fo± 0.3222±0.0045 QH

15 89.0 QV

16 89.0 QV 16 88.2 QV

17 157.0 QV 17 165.5 QH

18 183.31 ± 7 QH

19 183.31 ± 4.5 QH

19 183.31 ± 3 QH 20 183.31 ± 3 QH

20 191.31 QV 21 183.31 ± 1.8 QH

18 183.31 ± 1

QH 22 183.31 ± 1 QH

Exact match to AMSU/MHS

Only Polarization different Unique Passband Unique Passband, and Pol. different from closest AMSU/MHS channels

AMSU/MHS

6

MSU

Microwave Sounding Instruments from MSU to AMSU/MHS to ATMS

ATMS Field of View Size for the beam width of 2.2o – black line

ATMS Resample to the Field of View Size for the beam width of 3.3o- blue line

7

8

ATMS Channel Weighting Functions

Pres

sure

(hPa

)

Weighting Function

ATMS SDR Team 2015 Top Five Accomplishments

1. Developed the radiometric two-point calibration in radiance, instead of brightness temperature which is based on Rayleigh-Jeans approximation. The full radiance calibration algorithm will be in IDPS MX8.12 and IDPS Block 2

2. Standardized NEdT calculation for ATMS and other microwave sounding instruments using Allan Deviation. The new algorithm has resulted in much stable noise trending and is SI traceable

3. Optimized the ATMS de-striping algorithm for the earth scene brightness temperatures and generated 45 days of ATMS TDR data for NWP user community to experiment the impacts of ATMS on global forecast skills

4. Developed a physically based model for correcting the radiation from ATMS reflector emission contributed to the earth scene brightness temperature

5. Updated ATMS processing coefficient tables (e.g. nonlinearity coefficients, threshold for calibration counts)

9

S-NPP ATMS On-orbit Performance

10

Channel Accuracy (K) On-Orbit/Spec

NEΔT (K) On-Orbit/Spec

Channel Calibration (K) On-Orbit/Spec

NEΔT (K) On-Orbit/Spec

1 /1.00 0.25/0.5 12 0.24/0.75 0.59/1.0

2 /1.00 0.31/0.6 13 0.13/0.75 0.86/1.5

3 /0.75 0.37/0.7 14 0.02/0.75 1.23/2.2

4 /0.75 0.28/0.5 15 0.09/0.75 1.95/3.6

5 0.18/0.75 0.28/0.5 16 /1.00 0.29/0.3

6 0.09/0.75 0.29/0.5 17 /1.00 0.46/0.6

7 0.02/0.75 0.27/0.5 18 0.50/1.00 0.38/0.8

8 0.06/0.75 0.27/0.5 19 0.36/1.00 0.46/0.8

9 0.06/0.75 0.29/0.5 20 0.31/1.00 0.54/0.8

10 0.18/0.75 0.43/0.75 21 0.13/1.00 0.59/0.8

11 0.22/0.75 0.56/1.0 22 0.40/1.00 0.73/0.9

Note: On-orbit calibration accuracy for ATMS antenna brightness temperatures at upper air sounding channels is derived from the forward model (see Zou, X., Lin Lin and F. Weng, 2013: Absolute Calibration of ATMS Upper Level Temperature Sounding Channels Using GPS RO Observations , IEEE Trans. Geosci. and Remote Sens., 10.1109/TGRS.2013.2250981)

SNPP ATMS SDR CalVal Major Milestones

11

11/14/2013

Time 01/31/2013

Provisional status SDR uncertainties met

requirements Beta status

First IDPS SDR

Validated status Mx8.0

Lunar and striping correction

Mx8.4

03/18/2014

04/19/2012

04/02/2012

Full radiance calibration

Mx8.10

10/31/2015

ATMS SDR Science Advances

• Radiometric Calibration Non-linearity correction Calibration accuracy Lunar intrusion correction

• Noise Characterization Standard deviation Allan deviation

• SDR Algorithm TDR to SDR conversion Resampling SDR through Back-Gilbert theory Xcal with respect to AMSU for climate applications Striping and characterization

• Advanced Developments TDR correction from antenna emission Full radiance calibration

12

ATMS Noise Equivalent Temperature (NEDT)

13

• Define SI-traceable noise evaluation algorithm using Allan deviation method* • Channel noise by Allan deviation based algorithm is lower than that provided by heritage

standard deviation based algorithm • Annual oscillation of channel noise is removed • Long term trending of S-NPP ATMS channel noise by Allan deviation algorithm started to

be provided in STAR ICVS-LTM from June 17, 2015

Tian, M., X. Zou and F. Weng, "Use of Allan Deviation for Characterizing Satellite Microwave Sounders Noise Equivalent Differential Temperature (NEDT)", IEEE Geosci. Remote Sens. Lett., (Accepted).

Impacts of ATMS Striping Effects on Channel Noise Characterization

14

• Channel noise reduced after applying striping mitigation algorithm

• 45-day de-striping BUFR data generated for NWP impact study

Qin, Z., X. Zou and F. Weng, 2013: Analysis of ATMS and AMSU striping noise from their earth scene observations. J. Geophy. Res., 118, 13,214-13,229, doi: 10.1002/ 2013JD020399 Ma, Y. and X. Zou, 2015: Optimal filters for striping noise mitigation within ATMS calibration counts. IEEE Trans. Geo. Remote Sensing, (in revision)

ATMS Reflector Emission and Its Effects on TDR

• Flat rotating reflector has an emission and affects the accuracy in computing the calibration target temperatures in two point calibration equations

• In the earth scene scanning, the antenna brightness temperature in the two-point calibration equation contains emission that must be further corrected

• Hagen-Rubens equation

0.0025 to 0.0065

• An algorithm is being developed for

ATMS TDR correction 15

Gold Layer

Nickel interface

ATMS TDR Scan Bias from Pitch-Over Maneuver Data

Start maneuver 1815 UTC

B49 scan off earth view at 1826 UTC

B49 returns earth view at 1848 UTC

End pitch 1858 UTC

ATMS TDR at Ch18 on February 20, 2012 Channel 1 Channel 2

Channel 3 Channel 4

NPP ATMS pitch maneuver observations show channel related scan angle dependent feature, indicate the scan bias is not inherent feature of the scene

16

Effects of ATMS Plane Reflector Emission on Brightness Temperature

For Quasi-V (TDR) :

For Quasi-H (TDR)

22v h hε ε ε= −

At an incident angle of 45 degree to the plane reflector, the Fresnel equation becomes

The second and third terms are the biases related to the reflector emission

Yang, H. and F. Weng, 2015: Estimation of ATMS Antenna Emission from cold space observations, IEEE Geosci. Trans. Remote. Sens, Submitted

17

18

The Reflector-Emission Bias for Space View

19

The Reflector-Emission Bias for Earth Scenes

ATMS Full Radiance Calibration (FRC) Tested in ADL Environment

Package: ADL 4.2 with MX 8.8 Data Ingested: 6 orbits S-NPP RDR data (17829 – 17834 from GRAVITE) on April 7, 2015 Output Data:

• TDR/SDR/GEO using full radiance calibration (FRC) algorithm

Analysis Provided: • Global mean TDR-RTM bias (ADL-FRC vs IDPS) by channels based on 6

orbits data • Global mean TDR bias (ADL-FPC vs IDPS) by channels based on 6 orbits

data

20

Global Mean TDR-RTM Bias

-1

-0.5

0

0.5

1

1.5

2

2.5

3

3.5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

TDR

- RTM

Bia

s [K]

Channels

ATMS TDR-RTM Bias using FRP (Blue) and using IDPS OPS (Red)

TDR (FRP) - RTM TDR (IDPS) - RTM

ATMS full radiance calibration (FRC) performs two corrections: 1) replacing the brightness temperatures (R-J approximation) with Plank function radiance and 2) reversing the sign in nonlinearity term. WG bands are affected by two corrections where the rest bands are mainly affected by the nonlinearity term.

21

Global Mean TDR Bias

-0.7

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

-6E-16

0.1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Mea

n TD

R Bi

as [K

]

Channels

ATMS TDR Bias (Full Radiance Process - IDPS OPS)

TDR Bias (Full Radiance Process - IDPS)

22

JPSS-1 ATMS SDR Algorithm Readiness

Radiance calibration algorithm − A full radiance calibration is adopted as the standard calibration method for both the two-point linear

calibration and non-linear correction Nonlinearity correction algorithm consistent with NOAA/METOP AMSU-A/MHS − Maximum nonlinearity was expressed as a function of μ parameter − Nonlinear parameter was expressed as a function of instrument temperature Reflector emissivity correction − Physical model for antenna reflector emissivity correction Lunar intrusion correction algorithm − LI is modeled as a function of antenna response, solid angle of the Moon and the microwave emission from

the Moon − The new correction model with best fitted parameters from ATMS observations can effectively reduce the

calibration error due to lunar contamination on cold counts De-stripping algorithm − Based on power spectrum analysis, stripping index and de-striping algorithm was developed to reduce the

flicker noise in calibration data and TDR products − The flicker noise and correlation on the JPSS1 ATMS is much lower than S-NPP ATMS TDR Remapping algorithm − B-G algorithm was developed to explore the advantage of ATMS oversampling feature − By using B-G algorithm, remapping coefficients were generated offline, to remap ATMS observation to

different FOV size 23

ATMS SDR Algorithm Change from SNPP to JPSS

Input Radiometric (Scene, Warm Target, Cold Space) Counts, PRT Counts, Coefficients

Compute Warm Target Radiance

Correction of Warm Radiance

Compute Cold Space Radiance

Correction of Cold Radiance

Average over Warm and Cold Counts

Linear Calibration of Scene Radiance

Nonlinearity Correction

Scene Radiance/Brightness Temperature

Major Changes:

• Radiance based calibration

• Model based lunar contamination correction

• Updated parameterized nonlinearity correction

• Model based antenna reflector emissivity correction

Earth Scene Antenna Emissivity Correction

24

J1 New Code

J1 Major Changes

Repair

Existing Code

JPSS-1 Readiness • J1 Cal/Val Overview

o Beta Maturity: L+1 Month o Provisional Maturity: L+3 Months o Validated Maturity: L+12 Months o Pre-Launch Calibration/Validation Plans

Analyze J1 ATMS TVAC regression test data Derive coefficients for SDR algorithm and deliver JPSS-1 ATMS SDR PCT Test JPSS-1 proxy data for SDR algorithm functional testing Use JPSS-1 proxy data (from TVAC) to verify delivered PCT Analyze spectral response function datasets Verify instrument mounting matrix for geolocation accuracy assessment

o Post-Launch Calibration/Validation Plans Conduct 30+ post-launch cal/val activities following JPSS ATMS Cal/Val plan

25

JPSS-1 ATMS SDR Algorithm Tests with Proxy Data

• Proxy data – JPSS-1 ATMS RDR from S-NPP mission data – JPSS-1 ATMS RDR from JPSS-1 ATMS TVAC data – JPSS-1 ATMS spacecraft level RDR

• Test Results – Test results using JPSS-1 ATMS RDR from S-NPP mission data have been

compared with those from SNPP. IDPS code was updated to handle JPSS-1 granule ID (J01) Geolocation is not accurate. Updated data will be delivered for additional testing.

– JPSS-1 ATMS PCT will be verified using RDR from JPSS-1 ATMS TVAC data – Validation system readiness: The additional validation capabilities is currently being developed at STAR and

will be ready well before J1 launch.

26

JPSS-1 ATMS TVAC Data Analysis Prior to Rework (1/2)

SNPP TVAC Data (RC1 230K) J-1 TVAC Data (1/10/14)

27

JPSS-1 ATMS TVAC Data Analysis Prior to Rework (2/2)

Calib

ratio

n Ac

cura

cy (

K)

Channel Index Red – Calibration accuracy from nominal Thermal Vacuum (TVAC) data, Green – values obtained from the best TVAC data and Blue – specification

28

Summary

• JPSS-1 ATMS CalVal Plan has been developed, including task networks, role and responsibility, caval methodology, expected outcomes

• ATMS on-orbit NEDT is well characterized in new Allan deviation method. The performance meets specification

• ATMS house-keeping parameters are being monitored through ICVS to support NASA commanding operations of scan reversal.

• Antenna reflector emission is fully characterized and the algorithm for correcting the emission from the reflector is ready for implementation

• All the calval sciences are well documented and published through peer-reviewed process

29

Path Forward

• For Suomi NPP ATMS, we will continue refining the SDR processing system − Begin ATMS mission-cycle reprocessing − Closely monitor S-NPP ATMS health status after implementing scan drive daily reversal − Improve radiative transfer (RT) model for more accurate simulation of window channels and

cloud radiance measurements for validation − Refine SDR algorithm modules, including lunar correction, antenna emission, TDR to SDR

conversion at window channels, and de-striping algorithm

• For JPSS -1 ATMS, we continue supporting pre-launch testing, instrument characterization and calibration data development − Complete the analyze J1 ATMS TVAC regression data after rework − Characterize the ATMS side lobe and cross-pol from antenna pattern data − Study the impacts of J1 spectral response function on forward model

• For the JPSS polar follow-on mission − Support the waiver studies in future instruments − Support the new instrumentation

30